> For the complete documentation index, see [llms.txt](https://docs.nexthink.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.nexthink.com/understanding-key-data-platform-concepts/live-device-diagnostics.md).

# Live Device Diagnostics

Live Device Diagnostics give Nexthink AI agents live access to device data that goes beyond continuous telemetry, such as event logs, registry settings, peripheral status, and Wi-Fi connection quality. Combined with the device's history, Live Device Diagnostics let AI agents investigate issues that telemetry alone can't explain, without interrupting the employee.

* **Spark** uses Live Device Diagnostics when an employee asks for help with their own device.
* **The device troubleshooter in Workspace (beta)** uses Live Device Diagnostics when an IT administrator troubleshoots a device they select.

This page explains what Live Device Diagnostics are, what they cover, and how access is controlled.

### What Live Device Diagnostics are <a href="#user-content-what-collectors-are" id="user-content-what-collectors-are"></a>

Live Device Diagnostics, or diagnostics for short, are a set of individual, read-only diagnostics built into the Nexthink Collector. Each diagnostic reads one specific part of the device's current state, such as the connected cameras or the free space on each disk, and returns it in a structured format.

AI agents combine Live Device Diagnostics to troubleshoot an issue. They select the diagnostics relevant to the issue, the Collector runs them on the device, often several in parallel, and returns the results. AI agents interpret the results and decides which diagnostics to run next, or what to conclude.

Live Device Diagnostics complement the NQL data that Nexthink collects. NQL data shows how a device behaved over time. Live Device Diagnostics show its state right now. They run only when an AI agent requests them, so this data is not collected continuously from employees' devices.

### Live Device Diagnostics areas <a href="#user-content-what-collectors-can-check" id="user-content-what-collectors-can-check"></a>

All Live Device Diagnostics are **read-only**. They are organized by area, and an AI agent uses only the areas relevant to the issue.

Nexthink adds new areas and extends existing ones over time.

<table><thead><tr><th width="127.75">Area</th><th width="259.8388671875">What it covers</th><th width="268.310546875">Additional limits</th><th>Platform</th></tr></thead><tbody><tr><td>Hardware</td><td>Device model, firmware, processor, memory, and graphics card</td><td>—</td><td>Windows, macOS</td></tr><tr><td>Cameras</td><td>Connected cameras, which app is using the camera, and camera privacy settings</td><td>—</td><td>Windows</td></tr><tr><td>Audio</td><td>Playback and recording devices and default devices. On Windows, also volume and mute state and which apps are playing or recording sound</td><td>—</td><td>Windows, macOS</td></tr><tr><td>Displays</td><td>Connected monitors, graphics card, and resolution. On Windows, also refresh rate, scaling, and multi-monitor layout</td><td>—</td><td>Windows, macOS</td></tr><tr><td>Bluetooth</td><td>Bluetooth adapter status and paired devices with their connection state. On Windows, also remembered devices and their battery level</td><td>Lists only bluetooth devices the employee's device already knows and never scans for nearby devices</td><td>Windows, macOS</td></tr><tr><td>Other connected devices</td><td>USB, Thunderbolt, and other plug-and-play devices, with their status</td><td>—</td><td>Windows, macOS</td></tr><tr><td>Performance</td><td>Applications using the most processor, memory, graphics, and disk or network activity, running processes, Windows services, and performance counters</td><td>—</td><td>Windows</td></tr><tr><td>Storage</td><td>Disk capacity, free space, and disk health</td><td>—</td><td>Windows, macOS</td></tr><tr><td>User folder storage</td><td>Size of the employee's Documents, Downloads, and other personal folders, the largest files they contain, and the Recycle Bin</td><td>Returns file names and administrators can disable it. Refer to <a data-mention href="https://docs.nexthink.com/user-guide/spark/setting-up-and-managing-spark/managing-spark-settings.md#enabling-storage-troubleshooting-for-spark">Managing Spark settings and data inputs</a></td><td>Windows</td></tr><tr><td>OneDrive</td><td>OneDrive installation, sign-in, and synchronization status, and synchronization problems</td><td>—</td><td>Windows</td></tr><tr><td>Network</td><td>Network adapters, Wi-Fi connection quality, and IP addresses. On Windows, also DNS, gateways, and the routing table used to understand VPN behavior</td><td>—</td><td>Windows, macOS</td></tr><tr><td>Power</td><td>Battery charge and health, power source, and active power plan</td><td>—</td><td>Windows</td></tr><tr><td>Operating system and updates</td><td>Operating system version, uptime, and configuration. On Windows, also update history, pending updates, and whether a restart is required</td><td>—</td><td>Windows, macOS</td></tr><tr><td>Registry</td><td>Configuration keys and values stored in the Windows registry</td><td>Access to sensitive registry locations is blocked</td><td>Windows</td></tr><tr><td>Event logs</td><td>Windows event logs and the macOS unified log, filtered on the issue under investigation</td><td>Returns a limited number of recent entries per request. The Windows Security log is never read</td><td>Windows, macOS</td></tr><tr><td>Crash analysis</td><td>Analysis of Windows crash dump files after a system crash (blue screen)</td><td>Returns only the error code and the driver</td><td>Windows</td></tr></tbody></table>

### Data access and controls <a href="#user-content-data-access-and-controls" id="user-content-data-access-and-controls"></a>

Live Device Diagnostics apply the following controls every time an AI agent uses them.

* **Scoped to one device**: Live Device Diagnostics read only the device being troubleshot: the employee's own device in Spark, or the device the IT administrator selects in Workspace.
* **Run only when requested**: Live Device Diagnostics run only when an agent requests them. Nexthink does not collect this data continuously.
* **Read-only**: Live Device Diagnostics only read the device state.
* **Minimal permissions**: Live Device Diagnostics run with the permissions of the employee signed in to the device wherever possible, and keep system access to the minimum.
* **Limited access to sensitive sources**: The Windows Security log is never read, sensitive registry locations are blocked, and crash analysis returns technical details only.

### Data handling <a href="#user-content-data-handling" id="user-content-data-handling"></a>

Live Device Diagnostics return their results to the AI agent while it troubleshoots the issue. The full output of live device diagnostics is never stored in Nexthink Infinity.

The agent mentions only the results relevant to the issue in the conversation. For example, Spark applies file and path masking to protect privacy while keeping enough detail for conversation analysis. Spark displays only individual file names in a conversation, which IT administrators can see when monitoring a [Spark conversation](https://docs.nexthink.com/user-guide/spark/working-with-spark/monitoring-spark/monitoring-conversations.md). No other diagnostic data appears.

For more information on how Nexthink processes, stores, and protects data in its AI products, refer to [Global AI Hub](https://docs.nexthink.com/legal/global-ai-hub).


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://docs.nexthink.com/understanding-key-data-platform-concepts/live-device-diagnostics.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.
